folly::tape #2109
Closed
folly::tape #2109
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Summary: A better version of `vector<vector>` for some scenarios. Can also be an ok `vector<string>` if not everything is small vectors Basic usecase: building vector<vector> by pushing one element at a time, no reserve. ``` PushBackVecInts_20_0_15 1.28us 782.19K PushBackTapeInts_20_0_15 317.49ns 3.15M ``` Constructing a vector of strings (20 strings). When everything is SSO, there is no point in tape. However, when we allocate, tape begins to win big. ``` ConstructVecSmallStrings_20_0_15 122.31ns 8.18M ConstructTapeSmallStrings_20_0_15 129.04ns 7.75M ConstructVecLargeStrings_20_16_32 319.24ns 3.13M ConstructTapeLargeStrings_20_16_32 147.52ns 6.78M ``` Worst case scenario - 2 small strings. Overhead of tape makes this slower. We could tackle this but it requires very complicated code and so far, not the use case encountered. ``` ConstructVec2SmallStrings 20.95ns 47.74M ConstructTape2SmallStrings 38.83ns 25.75M ``` Differential Revision: D52260714
|
This pull request was exported from Phabricator. Differential Revision: D52260714 |
|
This pull request has been merged in b4ff536. |
facebook-github-bot
pushed a commit
to facebook/hhvm
that referenced
this pull request
Jan 2, 2024
Summary: X-link: facebook/folly#2109 A better version of `vector<vector>` for some scenarios. Can also be an ok `vector<string>` if not everything is small vectors ## Basic usecase 0: iteration for data, initially not in cache ``` IterateVecIntsCacheMiss_20_0_15 449.93ns 2.22M IterateTapeIntsCacheMiss_20_0_15 352.49ns 2.84M IterateVecStringsCacheMiss_20_0_15 262.35ns 3.81M IterateTapeStringsCacheMiss_20_0_15 222.20ns 4.50M ``` Tape is more cache friendly, very packed and predictable. ## Basic usecase 1: building vector<vector> by pushing one element at a time, no reserve. ``` PushBackVecInts_20_0_15 1.28us 782.19K PushBackTapeInts_20_0_15 317.49ns 3.15M ``` ## Worst case - constructing from a known range of SSO strings Constructing a vector of strings (20 strings). When everything is SSO, there is no point in tape. However, when we allocate, tape begins to win big. ``` ConstructVecSmallStrings_20_0_15 122.31ns 8.18M ConstructTapeSmallStrings_20_0_15 129.04ns 7.75M ConstructVecLargeStrings_20_16_32 319.24ns 3.13M ConstructTapeLargeStrings_20_16_32 147.52ns 6.78M ``` Worst case scenario - 2 small strings. Overhead of tape makes this slower. We could tackle this but it requires very complicated code and so far, not the use case encountered. ``` ConstructVec2SmallStrings 20.95ns 47.74M ConstructTape2SmallStrings 38.83ns 25.75M ``` Reviewed By: dmm-fb Differential Revision: D52260714 fbshipit-source-id: a90e01bc589646a38f2ca22c1b9589b6cd4c7f85
tekawade
reviewed
Jan 6, 2024
| : g(minLen + maxLen), len(minLen, maxLen) {} | ||
|
|
||
| const Cont& operator()() { | ||
| buf.resize(len(g)); |
There was a problem hiding this comment.
@DenisYaroshevskiy shouldn't this be buf.reserve(len(g)); instead?
resize + emplace_back() will increase the size. While reserve will reserve the space and array won't grow dynamically as often.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
A better version of
vector<vector>for some scenarios.Can also be an ok
vector<string>if not everything is small vectorsBasic usecase: building vector by pushing one element at a time, no reserve.
Constructing a vector of strings (20 strings). When everything is SSO, there is no point in tape. However, when we allocate, tape begins to win big.
Worst case scenario - 2 small strings. Overhead of tape makes this slower. We could tackle this but it requires very complicated code and so far, not the use case encountered.
Differential Revision: D52260714